The Rakai Community Cohort Study (RCCS) is an open cohort of ~15,800 persons, 15-49 in 50 communities in rural Rakai, Uganda (HIV prevalence ~12.3%, incidence ~1.0/100 py.) Using partnership-based transmission models, we estimate that 40% of annual transmissions are due to extra-household partners of whom ~62% are from known partners outside an individual's community of residence (i.e., extra-community contacts seed infections into the general population.) The sources of these infections are largely unknown. In 2011-13, surveillance was extended to key communities including fishing communities (HIV prevalence 35-43%, incidence ~3.9/100 py) and transport hubs. With PEPFAR support, the Program will scale up combination HIV prevention (CHP) using antiretroviral therapy, male circumcision, prevention of mother-to-child HIV transmission and demand creation/behavioral modification in Rakai District (population ~400,000). We will test the hypotheses that: i) Key communities seed infection into the general RCCS population, and ii) that CHP will reduce HIV incidence through changing transmission dynamics within and between the general rural and key populations. Aim 1 will assess transmission dynamics in and between general and key communities prior to CHP scale-up (2011-13). The spatial scale of sexual networks will be assessed using household GPS, location of partners and mobility. Phylogenetic analyses (Sanger sequencing of gag and env) will assess patterns, timing and directionality of HIV transmission. Transmission modeling based on sexual partner and phylogenetic data will quantify the proportions of transmissions occurring within and between general and key communities. This addresses hypothesis 1, and provides the counterfactual for CHP Impact Evaluation. Aim 2 will assess HIV incidence and transmission dynamics during scale up of CHP services (2013-17.) CHP coverage will be monitored. The hazard ratio of population-level HIV incidence associated with community CHP coverage and individual risk characteristics will quantify the impact of CHP. The study is powered to detect an expected 46% reduction in HIV incidence. Changes in transmission dynamics will be assessed by the methods for Aim 1. Results will inform CHP programs in rural Africa.